Being asked versus being told: A proposal to evaluate the effects of mandated environmental service-learning on environmental service
Beth Covitt
Introduction
In 1978, delegates to the United Nations sponsored environmental education (EE) meeting in Tblisi, Georgia proposed that the goal of EE should be to provide people with the awareness, knowledge, attitudes, skills, and motivations to solve environmental problems (Tblisi, 1978). This is a heavy burden to place on any field, but we (environmental educators) cannot complain too loudly, considering the fact that we have placed it on ourselves. We have defined our mission such that our students must not merely know about the environment, they must also act to protect and improve it. Whether or not we have succeeded in achieving this mission in the two plus decades since Tblisi is arguable, and I will not comment on this here further than to say that though much has been accomplished, much still remains to be done.
Through my studies and consideration of how to achieve the mission of EE, I have begun to focus on the sub-goal of providing people with "motivations." In particular, I have thought increasingly about the idea that students’ motivations may be undermined rather than fostered when extrinsic incentives such as rewards, requirements, punishments, or threats are used (Deci & Ryan, 1985). This idea, which is neither new (e.g., Deci, 1971) nor uncontroversial (e.g., Cameron & Pierce, 1996; Lepper, et al., 1996) in the field of psychology, has not really entered the body of knowledge and research being conducted in the field of EE. It does, however, raise a question pertinent to the mission of EE. In essence, when we require that students act on behalf of the environment, are we fostering students’ lasting commitments to take action on behalf of the environment, or are we undermining the intrinsic motives that students might develop towards the environment if they were encouraged rather than required to help?
In order to begin addressing this question in the context of EE, I propose to conduct a study contrasting the effects of mandated versus voluntary environmental service-learning on middle school students’ continuing commitment to engage in service. Service-learning may be defined as "the various pedagogies that link community service and academic study so that each strengthens the other" . Here, environmental service is defined as behavior intended to benefit the environment either directly (e.g., engaging in ecological restoration of a prairie ecosystem or starting a recycling program in a school) or indirectly (e.g., teaching other students about environmental issues or participating in local politics to help pass a parks millage).
This study is pertinent today both for environmental educators and more generally for educators in schools across the nations. Many school districts, and states are currently considering whether to require students to take courses with service components and to complete a certain number of service hours in order to graduate (Andersen, 1999). In fact, the state of Maryland has already made service a requirement for all students in its public schools (Sobus, 1995). It seems reasonable to pause before implementing such a requirement, to consider what the effects may be on students’ continuing commitment to serve after their mandated service hours have been fulfilled.
To address this issue, I plan to conduct a field experiment in middle school classrooms using an environmental service-learning curriculum. The study will seek to answer three primary research questions. First, is there an effect of mandatory versus voluntary service-learning on students’ continuing commitment to engage in service? Second, is there an effect of differing amounts of student self-determination in choosing or creating a service-learning project on students’ continuing commitment to engage in service. Finally, is there an interaction effect between mandatory versus voluntary and level of project self-determination conditions?
The third question is particularly interesting to me because it speaks to a practical question for environmental service-learning educators and policy-makers. The question addresses whether potential negative effects of mandated service can be alleviated by providing opportunities for meaningful choice, decision-making, and participation within students’ mandated service experiences. That is, if service-learning is mandated, are there ways to implement it so as not to undermine students’ motivations for continuing engagement in service?
Although the issue of mandated behavioral requirements has not been widely discussed in the field of EE, it has been hotly debated among those interested in service-learning. Proponents of mandated service-learning believe that mandated service will help foster civic engagement among students (Andersen, 1999). Furthermore, they have proposed that, "all students benefit from the lessons learned through service, and mandating service is the only way to reach those who would never participate voluntarily" (Andersen, 1999). Alternatively, opponents of mandatory service-learning cite evidence that creating external pressures or requirements for service behaviors might alienate some students from future service (e.g., Deci and Ryan, 1985; Sobus, 1995; Stukas, et al., 1999).
Review of literature pertinent to mandated service-learning
Little research specifically addresses the effects of mandatory versus voluntary service-learning on continuing service behaviors (e.g., Shumer, et al., 1999). Similarly, little research has been conducted on the topic of environmental service learning (e.g., Center for Human Resources, 1999). However, a great deal of work has been done on closely related topics that speak to the question of mandating service-learning. Examples of such work include studies of the effects of rewards on sustained interest (Cialdini, et al., 1998, Lepper, et al. 1973), of perceived control and autonomy on children’s behaviors and emotions in the academic domain (Patrick, et al., 1993); of controlling teaching strategies on children’s performance (Flink, et al., 1990), of promoting self-determination on students’ analytical problem-solving skills (Boggiano, et al., 1993), and of extrinsic incentives for helping on perceived altruism (Batson, et al., 1978). By considering the arguments for and against mandating service-learning, it is possible to highlight some of this pertinent research from a service-learning perspective, and consequently, to make some predictions concerning the effects of mandating service-learning.
First, the case against. Those who oppose mandatory service-learning cite a number of different reasons for their position. These include the arguments that service is "involuntary servitude" which violates the 13th Amendment (Andersen, 1999), and that it’s oxymoronic to teach the meaning of civic engagement in a democratic society by mandating service (Stanton, et. al, 1999). These are important arguments that speak to the essence of what it means to live in a democratic society. However, in my mind, these are philosophical arguments that do not necessarily speak to the practical EE mission of fostering environmentally responsible behaviors. Therefore, in this study, the emphasis will be placed on outcome based arguments for and against mandating service-learning. In other words, the arguments/theories included in this study must directly concern whether or not mandated service-learning is more or less likely than voluntary service-learning to foster continuing service by students, and must be amenable to empirical testing.
One theory that can be applied to the issue concerning mandated service-learning is cognitive evaluation theory (Deci & Ryan, 1985). In particular, the first proposition of cognitive evaluation theory relates to the importance of self-determination. It states that,
"[e]xternal events relevant to the initiation or regulation of behavior will affect a person’s intrinsic motivation to the extent that they influence the perceived locus of causality for that behavior. Events that promote a more external perceived locus of causality will undermine intrinsic motivation, whereas those that promote a more internal perceived locus of causality will enhance intrinsic motivation" (Deci & Ryan, 1985).
In the context of mandated service-learning, cognitive evaluation theory posits that because students are required to engage in service, they will attribute their motivation for serving to the requirement, rather than to an intrinsic desire to serve or an intrinsic sense of their own altruistic nature. Thus, when the requirement to serve is removed, students will not have integrated a commitment to environmental service into their self-perceptions of the kind of people they are. In other words, students who are required to serve, will be less likely to view themselves as individuals who would choose to serve of their own volition. Although no research has tested this theory as it relates to service learning, a number of studies have supported cognitive evaluation theory in the context of continuing motivations to engage in volunteer activities and a variety of learning tasks (e.g., Lepper, et al., 1973; Koestner, et al., 1984; Stukas, et al., 1999).
Next, the case in favor of mandating service-learning. One pro-mandate argument is that "mandating service is the only way to reach those who would never participate voluntarily" (Andersen, 1999). However, some evidence suggests that the number of people who would "never participate voluntarily" may be very small. In fact, the key to fostering service among a vast majority of Americans may be as simple as asking. A national Independent Sector (1996) survey found that respondents were four times more likely to volunteer if they were asked than if they were not asked. Furthermore, respondents who were black, Hispanic, or between the ages of 18 and 24 were less likely to be asked to volunteer. However, when asked, respondents in these groups were as likely, if not more likely to volunteer than the national average. When considered on a national scale, these results provide a very positive picture for the future of American service. In 1995, only 43% of all survey respondents had been asked to volunteer, and 85% of this group did volunteer. Of the 54% of respondents who had not been asked, only 21% actually volunteered.
In my mind, a more interesting argument in favor of mandated service-learning comes from those proponents who acknowledge that there may be negative effects to a mandate, but believe that these negative effects can be alleviated by what may be called "best practices" of service-learning. According to service-learning practitioners (e.g., Andersen, 1999), best practices for service-learning mean that (1) service-learning is integrated into a curriculum and involves the school and teachers in planning, (2) service-learning includes a reflective process of, for example, discussions and journal writing about service experiences, and (3) service-learning begins with an exploration process where students investigate their community, define a need and respond to it.
In essence, these proponents of mandated service-learning argue that the reason why previous studies of mandated service show detrimental effects (e.g., Stukas, et al., 1999) is because service in these studies did not follow best practices guidelines for service-learning. According to this argument, service-learning provides a significantly different experience from community service. Students engaged in service-learning will have more opportunities to integrate their service into their self-perceptions (i.e., through teacher involvement and support, structured opportunities to reflect on the personal meaning of their service, and learning about scientific, political, and social issues pertinent to their service in the classroom) (Wofford, 1999). The result of this increased integration is that students should perceive themselves as intrinsically service-oriented individuals.
By conducting a field study in classrooms where the "best practices" of service-learning are achieved, I propose to begin testing the de facto hypothesis that best practices for service-learning can alleviate the negative effects of mandated service. I will address the following research questions. First, is there an effect of mandatory versus voluntary service-learning on students’ continuing commitment to engage in service? Second, is there an effect of differing amounts of student self-determination in choosing and/or creating a service project on students’ continuing commitment to engage in service. Finally, is there an interaction effect between mandatory versus voluntary and level of project self-determination conditions?
This study will take place in middle school classrooms using environmental service-learning curricula specifically adapted to achieve the level of project self-determination conditions. An environmental service-learning context was chosen for this study in order to augment the amount of research about service-learning that is pertinent to the field of EE. In the Guidelines for the Initial Preparation of Environmental Educators recently published by the North American Association for Environmental Education (NAAEE, 2000), service-learning is included in the list of pedagogical methods suggested for teaching EE. However, little research has specifically supported the ability of service-learning to achieve the mission of EE (Shumer, et al., 1999).
The study will focus on middle school classrooms for several reasons. First, a disproportionate amount of EE in schools is conducted with middle school students, so this group is relevant for studying the effectiveness of a pedagogical method used in EE (Michalea, do you have a cite for this?). In addition, middle school students are often subject to school requirements such as mandated service, and are more limited in the courses they choose to take than, for example, high school students. Thus, the issue of self-determination is particularly salient for middle school students. Finally, because they are at an age where they begin to have a greater sense of people and environments outside of their immediate milieu (Fabes, et al., 1999), middle school students are an appropriate group to work with in considering an outwardly directed behavior such as environmental service .
Hypotheses
Figure 1. Predicted effects of conditions on continuing service
The prediction of an interaction effect (i.e., a steeper slope among the mandated as compared with the voluntary conditions) is based on the hypothesis that allowing for options and self-determination within mandated conditions may alleviate the negative consequences of mandating behavior. Although this hypothesis is consistent with Deci and Ryan’s (1985) cognitive evaluation theory, the idea that the effect of autonomy versus control works as a subjective continuum (i.e., that deleterious effects of control may be alleviated by providing some aspects of autonomy) has not been explicitly tested (Stukas, 2000). However, the hypothesis is implicit in how some researchers consider the topic. For example, in their discussion, Stukas, et al. (1999) mention that, "institutions that choose to impose community-service requirements may reduce inhibiting effects by giving students a sense of freedom and autonomy in meeting their requirements. Allowing participants to design the focus and specific details of their service may effectively solve this problem" (Stukas et al., 1999).
The prediction of increase in slope for both mandatory and voluntary conditions across the three project self-determination levels was based on the assumption that the creation process would foster higher levels of self-determination, depth of personal investment, and sense of project ownership, and that these factors, in turn, would lead to more positive attitudes towards service experience and greater continuing motivations to serve. In essence, this hypothesis revisits the idea that increasing self-determination will decrease negative effects of mandatory service.
However, it’s also possible that a potential lack of "knowing what to do" in the create condition as compared to the prefab conditions might make the creative condition more prone to project failure. Such failure would likely increase negative evaluations of the service experience and lead to a decrease in continuing service. Such an effect would work in the opposite direction from what was predicted in the hypothesis. (e.g., If some classrooms in the create condition fail to successfully launch a project, students’ perceived competence in engaging in environmental service may be reduced. If perceived competence is an important factor in determining levels of continuing service, then a positive effect for the create condition may not be observed.)
The prediction of the shallower slope among the voluntary as compared with the mandatory conditions is derived from minimal justification theory (Katzev, 1986). This theory posits that moderate (e.g., being asked) rather than powerful (e.g., being told) external justifications for targeted behaviors are more likely to lead to individuals viewing their future engagement in these behaviors as "internally, rather than externally, motivated" (Katzev, 1986). Thus, after initially volunteering to engage in service-learning, voluntary-no choice participants will have an initial sense of having autonomously committed to engaging in service, and therefore will not react negatively to lack of choice in project.
THE FOLLOWING SECTION IS UNDER CONSTRUCTION!!
(Not that the other sections aren’t, but this one is more so)
Methods
Overview: This study will consist of a field experiment in 7th and 8th grade middle school classrooms using a Chesapeake Bay Foundation environmental service-learning curriculum which will be specifically altered for the study. The curriculum will consist of two units. The first unit will cover knowledge about ecological, social, and political issues concerning the Chesapeake Bay and its environs. The second unit will comprise the service component of the curriculum, and will differ for each of the following six experimental conditions:
In addition, a group of control classrooms will use only the knowledge component of the curriculum. The curriculum will be employed over the course of one semester.
Discussion of sampling and design issues:
Participants will be 7th and 8th grade students attending schools in Maryland and Virginia. Maryland is a state where service is a mandatory requirement for students (Sobus, 1995). In Virginia, service is not a state-mandated requirement. Based on the state dichotomy, random assignment will not be possible for this study (i.e., it will take the form of a quasi-experiment). Because random assignment will not be achieved, it will be important to gather additional data to clarify any differences there may be between schools in Maryland and schools in Virginia. Thus, for example, data including standardized test scores, socioeconomic status, student/teacher ratios, and school enrollments will be collected. Attempts will be made to match schools in the two states on some significant characteristics.
Random assignment may be used for the second factor in the study. However, random assignment for the second condition will be contingent on the willingness of teachers to teach any of the three types of service-learning curricula. If, based on a preliminary investigation, it seems likely that a significant number of teachers will not be amenable to being assigned to a condition, then we will have to allow some to choose the condition for their classroom(s). Because there may be systematic differences in teachers’ willingness to be assigned to conditions (e.g., younger teachers may have had more pre-service professional development involving service-learning, and thus may be more willing to take on the "create" as opposed to the "prefab" project conditions), we will need to collect data on teachers and try to assure that we have sufficient numbers of types of teachers (e.g., varying ages, distributions of gender, varying experience with teaching environmental education and service-learning) in all cells.
Also, it is unlikely that most teachers will be willing to use three different curricula with their different classes. Many teachers may not even teach three classes that would be appropriate for this study. Thus, realistically, there’s very little chance that we could achieve a sample where all schools or teachers would have classrooms in all three self-determination level conditions. This issue will have to be handled so as to minimize problems caused by lack of randomization. Where possible, attempts will be made to achieve homogeneity or at least similar distributions across cells on additional variables of possible import such as student ability,
A hierarchical linear modeling approach would provide a more precise tool for partitioning error terms in our study. Thus, if possible, we would like to include a sufficient sample to use this tool. Dr. Lepkowski suggested an educated guess that 20 classrooms per condition might be sufficient. He also said he would be willing to help me with more precise calculations as the project progresses.
Procedure
Teacher training
Teachers will attend two-day training sessions to learn how to implement the service-learning curriculum, how to administer pre and post tests, and how to complete a teacher log. . Training sessions will be conducted separately for each of the four treatment groups, and will reflect the conditions of the specific group. In addition, ongoing support will be provided to teachers throughout the study by project researchers and Chesapeake Bay Foundation staff.
Pretest
Pretests will be administered to all students to determine pre-treatment levels of a variety of variables. One measure of particular import will be pre-treatment level of (environmental) service (e.g. in hours/month). By collecting these data, it will be possible to use a "change in level of service" measure as the dependent variable. Other variables that may be important to include are environmental sensitivity, past (environmental) service experiences, perception of teacher’s autonomy support, attitudes toward service (including extent to which students feel they would serve only if required to do so), preliminary intent to engage in environmental service, age, gender, religious affiliation, and religious beliefs about service. Other important information to collect about students may include socioeconomic status, and parents’ professions. Teacher pretests will also be administered to collect data including teacher’s age, gender, level of training and experience with environmental education and service-learning, and attitudes toward environmental education and service-learning.
Mandatory/voluntary induction
Induction for Maryland students will emphasize that service is a state requirement for all students. Induction for Virginia students will make service voluntary. Students will be told that they may choose to do an independent issue investigation project rather than a service project. However, an induced-compliance technique will be used to encourage all students to participate in the service project (Stukas, et al., 1999). The issue investigation project will be designed and presented so that students understand that it will require roughly the same amount of work and effort as the service-learning project.
Choice in project/No choice in project induction
Next, according to level of self-determination in project condition, students will either be told (1) that the class will be engaging in a specific project (e.g., growing submerged aquatic vegetation and planting it in the Chesapeake Bay), (2) that the class will decide on one of three projects (e.g., submerged aquatic vegetation, restoring an Oyster bed to the Bay, or raising and releasing shad in Chesapeake Bay tributaries), or (3) that the class will conduct a community inventory and decide upon, develop, and implement an environmental service project.
Due to logistical constraints on middle school teachers, each classroom will engage in only one service project. This means that even in the two conditions where students have options about projects, some students will presumably not be doing their "first choice" projects. However, because the decision will be made through a democratic process with classmates, rather than as an authoritarian mandate from the teacher, perception of autonomy should hopefully still be greater in these conditions than in the no option condition. (I need to find evidence for this).
Service-learning curriculum implementation
Classes will engage in respective service-learning curricula/projects over the course of approximately one semester. During the first part of the semester, classrooms will complete the knowledge unit about the Chesapeake Bay. During the second portion of the semester, the classrooms will engage in service-learning projects. Efforts will be made (through curriculum specifications and teacher training sessions) to make some potentially important additional variables homogeneous across all six groups. For example, curricula and training sessions will emphasize to teachers in all groups the importance of the reflection process in service-learning.
Throughout the period of the curriculum implementation (i.e., after completion of the knowledge component, early in the service-learning project, later in the service learning project) data will be collected about students intentions to continue serving after the project is over. In this way, a shape of response over time can be developed to have a better sense of how the various phases of the curriculum are affecting the students.
In addition, student absenteeism during the study will be recorded because absenteeism may influence students’ experiences with and their continuing commitment to serve.
Immediate Post-test (Appendix A)
Immediately following completion of the service-learning unit, post-tests will be administered to students. Students’ post-tests will likely include items to measure perceived quality of service experience, attitudes towards the environment and towards service, and intentions to continue (environmental) service.
Opportunities to serve
Through the Chesapeake Bay Foundation student alumni organization, all students in the study will be provided with opportunities to continue engaging in environmental service. These opportunities may, for example, take the form of weekend workdays and Megalops newsletters with announcements and suggestions of "things you can do to help the Bay." Efforts will be made to help all interested students participate (e.g., considerations will be made for funding and transportation). By providing these opportunities, it should be possible to achieve a reasonable range in continuing service by students.
Delayed post-tests
Additional post-tests (three months, six months, one year after completion of course) will be administered to students to measure their actual continuing (environmental) service and their continuing intentions to serve in the future.
Models
Below is the "bare-bones" ANOVA model. Another possibility would be to run the models as regressions.
Mandatory/Voluntary x No Option Prefab/3 Options Prefab/Opt & Create ANOVA Model
Yijk = m + a i + b j + a b i j + e ijk
WhereY is the individual’s level of continuing service, i refers to level of treatment A (mandatory/voluntary), j refers to level of treatment B (no option prefab, 3 options prefab, opt & create), and k refers to the subjects in the i,j treatment combinations. The assumptions for this model are equality of cell variances, normality of cell distributions, and independence. However, because of nonrandom assignment, independence is not achieved.
Because this will be a quasi-experiment with classroom and school clustering issues, it will probably be important to measure some additional variables that might otherwise be accounted for by random assignment. By measuring such variables I should be able to include them as covariates in the model to determine whether their inclusion reduces/eliminates any effects I might find for the primary independent variables. However, I’m still in the process of deciding which "additional variables" I should consider, so for this draft I decided to keep my models to the simplest case.
Addressing unequal cell sizes due to treatment effect
Based on the vagaries of class sizes, it seems foreordained that my treatment groups will not consist of equal numbers of participants. Although this fact in itself does not proscribe the use of ANOVA (i.e., it’s possible to use unique approach to decomposing main effect terms), there is another factor likely to affect cell sizes in a systematic manner. The mandatory/voluntary dichotomy raises the possibility of participant attrition in the voluntary condition due to choice. Based on previous evidence, however, there are several potential means of handling this issue. First, the Independent Sector (1996) survey suggests that most individuals do serve when asked. Also, in their study, Stukas, et al. (1999) were able to avoid this problem by using an induced-compliance technique (i.e., participants in voluntary condition were told that they did not have to volunteer, but that "it would really help if they did"). All participants in their voluntary group agreed to engage in service.
Another way to handle the problem of systematic attrition in cell size would be to analyze the whole cell in the voluntary condition (i.e., both those who opt to serve and those who opt out). In this way, an overall effect of the "voluntary" treatment can be observed, however, the effect of "doing service" won’t be separated. It’s likely that by including the whole cell in the analysis, any effect in the hypothesized direction in comparison with the mandatory condition will be diluted. However, it would also be possible to analyze the data both with the whole voluntary cells and with the reduced voluntary cells. In this way, a sort of bounds for the effect may be determined.
Conclusion
There are many questions that still need to be answered about the effectiveness of service-learning for promoting EE citizenship goals. Within the context of my dissertation research, it will not be possible to answer all the questions that need to be addressed. However, by working toward answering several key questions about mandatory environmental service, my research will increase our knowledge of how to implement service-learning courses to achieve the most positive impact on students’ commitments to continuing service. This knowledge will contribute both to the EE mission of fostering environmental problem-solving behaviors, as well as to the more general educational goal of fostering positive civic engagement among students.
(References available from author)